Epicardial fat can predict the in-atrial thrombus in patients with atrial fibrillation

Abstract Background Thromboprophylaxis is important in the management of patients with atrial fibrillation. Epicardial adipose tissue (EAT) is involved in the formation of the pre-thrombotic state. However, the potential of EAT in predicting the risk of thromboembolism in atrial fibrillation remains unknown. Objective To study the imaging features of EAT and its potential in predicting the risk of in-atrial thrombus in atrial fibrillation. Methods Patients were divided into the group with (thrombus group) and without (non-thrombus group) intra-atrial thrombosis according to the results of transesophageal echocardiography (TEE). The EAT volume, mean EAT density and EAT volume/whole heart volume ratio measured on coronary artery CT angiography (CA-CTA) were compared between the two groups. Result A total of 347 atrial fibrillation patients undergoing both TEE and CA-CTA were included and 38 cases were complicated with in-atrial thrombus. Using quartile values, EAT volume, mean EAT density, the ratio of EAT volume/whole heart volume are divided into four grades according to the values. By analyzing the ROC curve and comparing the area under the curve of AUC, EAT volume and EAT volume/whole heart volume ratio show potential diagnostic value in predicting the risk of in-atrial thrombus in atrial fibrillation. In addition, 38 cases without in-atrial thrombus were included in the non-thrombotic group after matching the propensity score by age and gender. On multivariate analysis, patients in the thrombus group had the larger EAT volume (P=0.000, P<0.05) and higher ratio of EAT volume/whole heart volume ratio (P=0.004, P<0.05) when compared with patients in the non-thrombotic group, but there was no difference of mean EAT density (P=0.224, P>0.05) between the two groups. On logistic regression analysis, patients in the thrombus group had the larger EAT volume when compared with patients in the non-thrombotic group (P=0.003, P<0.05). Conclusions EAT may be a potential marker in predicting the risk of in-atrial thrombus in atrial fibrillation. However, much more efforts and testing would be needed before our conclusion could be tried in other people and to elucidate pathogenic mechanisms of EAT.Figure 1Epicardial fat measuremeFigure 2ROC curve